Gliding Ability of the Siberian Flying Squirrel Pteromys Volans Orii

Mammal Study 32: 151–154 (2007) © the Mammalogical Society of Japan

Gliding ability of the Siberian flying squirrel Pteromys volans orii

Yushin Asari1,2, Hisashi Yanagawa2,* and Tatsuo Oshida2 1 The United Graduate School of Agricultural Science, Iwate University, Morioka 020-8550, Japan 2 Laboratory of Wildlife Ecology, Obihiro University of Agriculture and Veterinary Medicine, Obihiro 080-8555, Japan

Abstract. Forest fragmentation is a threat to flying squirrel population due to dependence on gliding locomotion in forests. Therefore, it is essential to understand their gliding ability. The gliding locomotion of Pteromys volans orii, were observed from July 2003 to June 2005, in Obihiro, Hokkaido, Japan. The horizontal distance and glide ratio obtained from 31 glides were employed as indicators to know their gliding ability. The gliding ability was not affected by weight and sex in the Siberian flying squirrel. Mean horizontal distance and glide ratio were 18.90 m and 1.70 with great variation. Although maximum values were 49.40 m (horizontal distance) and 3.31 (glide ratio), most of the horizontal distance and glide ratio were in the ‘10–20 m’ and ‘1.0–1.5’, respectively. Therefore, to retain the flying squirrel populations, forest gaps should not exceed the distance traversable with a glide ratio of 1.0 (distance between forests/tree height at the forest edg).

Key words: gliding ability, gliding ratio, horizontal distance, Pteromys volans orii.

The Siberian flying squirrel, Pteromys volans, ranges The key purposes of gliding may be to avoid predators over northern parts of the Eurasia, and from Korea to or to minimize the energy costs of moving (Goldingay Hokkaido, Japan. This species is a nocturnal gliding 2000). Gliding locomotion has been investigated in mammal. The flying squirrel population in Finland has some mammalian species (Ando and Shiraishi 1993; declined due to a shrinkage of the area of its optimal Jackson 1999; Vernes 2001). Flying squirrels are occa- habitat, which comprises stands of mature spruce and sionally exposed to high risks such as strong wind and broadleaved trees (Hokkanen et al. 1982). P. volans predators, when they travel across wide gaps between orii which is endemic subspecies to Hokkaido is widely fragmented forests. It is of particular importance to iden- distributed in the lowlands and mountains. Therefore, tify what is the secure gliding environment for flying it is not listed as a threatened species in Japan (Ministry squirrels in forests. Therefore, we observed the gliding of the Environment 2002). However, its forest habitats locomotion of P. volans orii, and estimated their gliding are shrinking due to land development. Residential areas abilities. and road construction have reduced or fragmented the forests. Materials and methods Unlike terrestrial mammals, flying squirrels depend on gliding locomotion. Although gliding is an effective Study area locomotion mode in forests, it would be difficult for The study was conducted in Obihiro, Hokkaido, Japan them to travel across wide-open area. Thus, forest frag- (42°51'–53'N, 143°9'–11'E). This city has a large resi- mentation could be a threat to populations of P. volans dential area and extensive agricultural land which pro- orii. Yanagawa et al. (2004) constructed a structure to vides habitats for flying squirrels. Some habitats are cross roads to mitigate the problems caused by forest fragmented and isolated, without tree connectivity. Our fragmentation in Hokkaido. Although there has been study areas of eight forests (1.7–3.1 ha) were located some action to conserve P. volans orii, its biological near the town, in strips along riverbanks, and in built-up traits are not fully understood. areas and meadows (Fig. 1). These areas are flat and

*To whom correspondence should be addressed. E-mail: [email protected] 152 Mammal Study 32 (2007)

Fig. 2. Gliding path of the flying squirrel and four measurements: 1) height of launch, 2) height of landing, 3) horizontal distance, and 4) vertical drop.

we observed their gliding behavior. Most animals first climbed the tree in which the nest box was installed, then glided from a high point of this tree to a lower point of another tree. To avoid disturbing them, we observed their gliding locomotion without the use of artificial light when possible. Observations were not carried out on Fig. 1. Landscapes of our study area. (A) shows a small forest adja- gusty or rainy days due to the impact of these conditions cent to roads, (B) shows a forest on the bank of a river. on the glide path. approachable, and are occupied by broadleaf trees such Measurement of glides as Betula platyphylla var. japonica, Alnus japonica, and Based on the observations of gliding locomotion, the Quercus dentata, and conifers such as Abies sachalinen- height from ground to launch point, height from ground sis and Larix leptolepis. The mean tree height in the for- to landing point, and horizontal distance between launch ests was 11.5 to 19.1 m, and the mean density was 116.7 and landing trees were recorded. Vertical drop was cal- to 908.3 trees/ha. There were sparse branches in the culated as the difference between height of launch and mid-story and shrubs in the understory, therefore it was height of landing (Vernes 2001) (Fig. 2). Each was mea- considered as the environment with few obstructions for sured by laser rangefinder (Opti-Logic Corporation) to gliding by flying squirrels. the nearest 0.1 m. Horizontal distance and glide ratio (horizontal distance/vertical drop) were employed as Field observations indicators of gliding ability (Ando and Shiraishi 1993). More than 100 nest boxes were spread throughout the study areas at 2 to 5 m height, as reported by Yanagawa Results (1994). From July 2003 to June 2005, 64 nest boxes were checked more than twice a month (excluding A total of 69 flying squirrels (35 males and 34 December to February) in the daytime. When the resting females) were captured using the nest boxes. Excluding flying squirrels were found, these boxes were collected young flying squirrels, 31 glides (18 males and 13 to capture them. Body mass, sex, reproductive condi- females) were recorded. Same individuals were included tion, and age class were recorded for all captured flying in all 31 glides. Most observed glides were the first squirrels. All animals were marked with numbered glides after leaving the nest box, but 9 glides were made metal ear tags. Nest boxes with marked individuals were by four squirrels on successive travels. Glides were returned to the capture point before sunset. After sunset, measured easily, since most flying squirrels traveled in a Asari et al., Gliding ability of Pteromys volans orii 153

Table 1. Parameters of gliding and weight for males and females

Males (n = 18) Females (n = 13) Statistic Parameters Mean SD Min Max Mean SD Min Max UP Height of launch (m) 15.71 4.32 8.50 23.80 13.45 5.44 5.60 24.00 86.50 ns Height of landing (m) 3.78 2.29 1.00 8.60 2.67 1.67 0.80 6.70 81.00 ns Vertical drop (m) 11.92 3.97 6.70 18.60 10.78 5.49 3.80 22.20 95.50 ns Horizontal distance (m) 20.62 11.82 4.30 49.40 16.52 9.45 5.70 39.00 90.50 ns Glide ratio 1.70 0.75 0.61 3.31 1.69 0.75 0.48 2.97 115.00 ns Weight (g) 110.20 10.79 91.30 123.30 132.18 28.67 73.70 161.40 47.00 <0.01

Table 2. Mean and range of the gliding performance for all glides

All glides (n = 31) Parameters Mean SD Min Max Height of launch (m) 14.76 4.87 5.60 24.00 Height of landing (m) 3.32 2.10 0.80 8.60 Vertical drop (m) 11.44 4.62 3.80 22.20 Horizontal distance (m) 18.90 3.97 4.30 49.40 Glide ratio 1.70 0.74 0.48 3.31 straight line, without turning, and the topography was Fig. 3. Observed numbers of horizontal distance divided by four flat. There were no significant differences in the gliding classes. parameters between male and female (t-test, P > 0.05; Table 1), although females (mean ± SD; 132.18 ± 28.67 g) were significantly heavier than males (110.20 ± 10.79 g; P < 0.01). The data for males and females were there- fore pooled in the following analysis. Moreover, body weight did not correlate with either horizontal distance or glide ratio (Pearson’s correlation coefficient; P > 0.05). The mean horizontal distance and glide ratio were 18.90 m and 1.70, respectively (Table 2). However, there was a wide range of horizontal distances (4.30– 49.40 m) and of glide ratios (0.48–3.31). The various glides we observed allowed us to identify the most com- Fig. 4. Observed numbers of gliding ratio divided by six classes. monly used type of glide. To determine the frequency distribution of the glides, we categorized them into 4 classes of horizontal distance and 6 classes of glide ratio. differences in gliding indicators between male and The most frequently traveled distance was 10 to 20 m (17 female. They had excellent gliding ability; that was 50 cases). The other distance classes were rarely observed m for horizontal distance and 3.3 for glide ratio in maxi- (Fig. 3). The most frequent glide ratio was in the ‘1.0– mum. However, the most observed glides of this species 1.5’ class (10 cases) (Fig. 4). The ‘≥3.0’ class had only were 10 to 20 m for their horizontal distance and 1.0–1.5 one case, but the other classes had five cases each. In for glide ratio. This may be to avoid high risks such as a addition, horizontal distance was significantly positively strong wind and avian predators in crossing wide gaps. correlated with glide ratio (r = 0.63, P < 0.0001). The mean horizontal distance of the Siberian flying squirrel was similar to that of the northern flying squir- Discussion rel, Glaucomys sabrinus (14.2–19.0 m; Vernes 2001), which has a similar weight (Goldingay 2000). Similarly, In Siberian flying squirrels, there were no significant there was little difference in the mean glide ratios 154 Mammal Study 32 (2007) between Siberian flying squirrels and northern flying flying squirrels may be high in areas where distances squirrels (1.98; Vernes 2001). Also, the sugar glider, between patches is <500 m (Woodworth et al. 2000). Petaurus breviceps, which has the same body size as the The effects of clear-cuttings negatively affect the biology Siberian flying squirrel, glides as flying squirrels (20.42 and occurrence of the flying squirrels, so the squirrel’s m; Jackson 1999). The southern flying squirrel, G. need to glide should be supported. If clearings exceed volans, which is smaller than the Siberian flying squir- the gliding ability of flying squirrels, developments must rel, has a mean glide ratio of 1.53 (Scheibe and Robins include measures to permit movement among forests 1998), but a larger species, Japanese giant flying squirrel (e.g., Eco-bridge for flying squirrels, Yanagawa et al. (Petaurista leucogenys), has glide ratios of 1 to 3 (Ando 2004). and Shiraishi 1993). It would appear that smaller species have lower glide ratios. Our data show, however, that Acknowledgements: We thank students of the Labora- there was no significant difference between squirrels of tory of Wildlife Ecology, Obihiro University, for helping different weights in a same species. Gliding ability of in our field surveys. We also thank Miss C. Nakano, the Siberian flying squirrel was also unaffected by their who drew the illustrations of the flying squirrel. weights. The Siberian flying squirrel showed a wide range of References horizontal distances and glide ratios, with the maximum for both being somewhat higher than that seen in other Ando, M. and Shiraishi, S. 1993. Gliding flight in the Japanese giant gliding mammals (G. sabrinus, Vernes 2001; Petaurus flying squirrel Petaurista leucogenys. Journal of Mammalogical breviceps, Jackson 1999; Petaurista leucogenys, Ando Society of Japan 18: 19–32. Goldingay, R. L. 2000. Gliding mammals of the world: diversity and and Shiraishi 1993). We therefore suggest that the Sibe- ecological requirements. In (R. Goldingay and J. Scheibe, eds.) rian flying squirrel also has a high gliding ability, match- Biology of Gliding Mammals. Pp. 9–44. Filander Verlag, Fürth. ing that of other gliding mammals. This higher gliding Hokkanen, H., Törmälä, T. and Vuorinen, H. 1982. Decline of flying ability could often be observed at widely fragmented squirrel Pteromys volans L. populations in Finland. Biological Conservation 23: 273–284. forest edges, where the gliding ratio was positively Jackson, S. M. 1999. Glide angle in the genus Petaurus and a review correlated with horizontal distance. In general, how- of gliding in mammals. Mammal Review 30: 9–30. ever, flying squirrels would be prevented from traveling Ministry of the Environment. 2002. Threatened Wildlife of Japan –Red Data Book 2nd ed., Vol. 1, Mammalia. Japan Wildlife among fragmented forests without an intermediate tree Research Center, Tokyo, 177 pp. (in Japanese). due to the limits on their practicable gliding distance. Scheibe, J. S. and Robins, J. H. 1998. Morphological and performance The maximum horizontal distance of the Siberian attributes of gliding mammals. In (M. A. Steel, J. F. Merritt and flying squirrel may be 50 m with a glide ratio of 3.3. D. A. Zegers, eds.) Ecology and Evolutionary Biology of Tree Squirrels. Pp. 131–144. Virginia Museum of Natural History, Nevertheless, most of its horizontal distances and glide Virginia. ratios were up to 20 m and 3.0, respectively. In addition, Selonen, V. and Hanski, I. K. 2004. Young flying squirrels (Pteromys most glides were in the ‘10–20 m’ and ‘1.0–1.5’ classes. volans) dispersing in fragmented forests. Behavioral Ecology 15: Although they have a comparatively high gliding ability, 564–571. Vernes, K. 2001. Gliding performance of the northern flying squirrel they generally travel shorter distances. The reason for (Glaucomys sabrinus) in mature mixed forest of eastern Canada. this conservatism might be the psychological and techni- Journal of Mammalogy 82: 1026–1033. cal difficulties seen in Petaurista leucogenys (Ando and Woodworth, C. J., Bollinger, E. K. and Nelson, T. A. 2000. The effects of forest fragment size, isolation, and microhabitat vari- Shiraishi 1993). Therefore, when their forest habitat is ables on nest box use by southern flying squirrels (Glaucomys fragmented by development, the gaps between forests volans) in southern Illinois. In (R. L. Goldingay and J. Scheibe, should be less than the distance traversable with a glide eds.) Biology of Gliding Mammals. Pp. 135–147. Filander ratio of at least 1.0 (= distance between forests/tree Verlag, Fürth. Yanagawa, H. 1994. Field observation of flying squirrels using nest height at the forest edge) by adult flying squirrels. On boxes. Forest Protection 241: 20–22 (in Japanese). account of the reduced abilities of young flying squirrels, Yanagawa, H., Asari, Y., Kishida, K., Kimura, S. and Kitase, T. 2004. lower distances and glide ratios should be considered. Eco-bridge for flying squirrels. Symposium on Wildlife and In Finland, Selonen and Hanski (2004) showed that Traffic Proceedings 3: 13–18 (in Japanese with English sum- mary). wide open areas prevent the straight movement of young Siberian flying squirrels. Also, the density of southern Received 17 December 2006. Accepted 20 May 2007.